Recent advances in Fourier analysis have brought new tools to efficiently represent and learn set functions. In this paper, we bring the power of Fourier analysis to the design of combinatorial auctions (CAs). The key idea is to approximate bidders' value functions using Fourier-sparse set functions, which can be computed using a relatively small number of queries. Since this number is still too large for practical CAs, we propose a new hybrid design: we first use neural networks (NNs) to learn bidders' values and then apply Fourier analysis to the learned representations. On a technical level, we formulate a Fourier transform-based winner determination problem and derive its mixed integer program formulation. Based on this, we devise an iterative CA that asks Fourier-based queries. We experimentally show that our hybrid ICA achieves higher efficiency than prior auction designs, leads to a fairer distribution of social welfare, and significantly reduces runtime. With this paper, we are the first to leverage Fourier analysis in CA design and lay the foundation for future work in this area. Our code is available on GitHub: https://github.com/marketdesignresearch/FA-based-ICAs.
翻译:Fourier分析的最新进展带来了高效地代表并学习设定功能的新工具。 在本文中,我们把Fourier分析的力量带到组合拍卖的设计中。 关键的想法是使用Fourier-sparse成套功能来估计投标人的价值功能, 使用相对较少的查询量来计算这些功能。 由于这一数字对于实际的CA来说仍然太大, 我们建议一个新的混合设计: 我们首先使用神经网络来学习投标人的价值, 然后对学习的演示应用Fourier分析。 在技术层面, 我们提出了一个基于Fourier变异的赢家的确定问题, 并得出其混合整数程序配置。 在此基础上, 我们设计了一个反复的 CA, 询问基于 Fourier 的查询 。 我们实验性地显示我们的混合ICA 取得了比先前的拍卖设计更高的效率, 导致社会福利的更公平的分配, 并大大缩短运行时间 。 有了这个文件, 我们首先在 CAA 设计中利用 Fourier 分析, 并为未来这一领域的工作打下基础。 我们的代码可以在 GitHub: https://github.com/markeddedesign deskendrestratreas- fas- fas- as- as- amakeatbs.